[
https://issues.apache.org/jira/browse/CRAIL-27?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16520100#comment-16520100
]
Jonas Pfefferle commented on CRAIL-27:
--------------------------------------
Yes but neither FileSystem or AbstractFileSystem implement all statistics, e.g.
you have to adapt stats in input and output streams. I did this analogous to
HDFS's implementation for Crail here:
[https://github.com/PepperJo/incubator-crail/commit/772355de779793d8675624dfe574e1c553072698]
Also if you use the delegation filesystem you don't need to implement it twice.
Take a look at:
[https://github.com/PepperJo/incubator-crail/commit/dfb1b211cf0a6427132eac3933e8a2bf935c4440]
I can make a pull request with these changes.
> Do not reimplement HDFS adapter for AbstractFileSystem
> ------------------------------------------------------
>
> Key: CRAIL-27
> URL: https://issues.apache.org/jira/browse/CRAIL-27
> Project: Apache Crail
> Issue Type: Improvement
> Reporter: Jonas Pfefferle
> Priority: Minor
>
> Use DelegateToFileSystem class to implement AbstractFileSystem instead of
> reimplementing all the calls.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
